Aim

Part of Deep Processing, the parent dimension for encoding work.

Aim is the list of questions a topic must answer, written before the material is opened: why each concept matters, and how the concepts relate. The questions start a search. Answering them means going through the material looking for importance and for relationships, and that searching is what does the encoding.

Aim is the first pass of the Bear Hunter System, the three-pass encoding loop inside Deep Processing. The three passes are Aim, Shoot, and Skin. Shoot works the material into a map. Skin prunes the map down to what survives without the material. Written first, the questions change what the reading does. Material arrives into a standing frame, already being compared and placed, instead of going past unconnected. Everything read afterwards has somewhere to land.

Core takeaways

• Two questions carry the whole pass: why is this important, and how does it relate to that other concept. Neither asks what something is, because a definition learned on its own is mostly gone within minutes.
• Thirteen skills have to be in place before the first Aim, and three of them are drills the step is built from: the traffic light drill, order control, and biggest chunks first. A missing skill means more practice, and never a smaller Aim.
• The keywords come off a fast skim of every resource, out of the material’s order. A topic holds 15 to 30 major concepts. About 80 percent of them in about 25 minutes beats all of them in two hours.
• The two questions go over the list, the concepts group into rough chunks, and the pass ends with a backbone of the biggest chunks and no details on it. The structure is held loosely, because effort already spent on a structure makes it feel too expensive to break.
• The pass is meant to feel chaotic, confusing, overwhelming, and slow. A pass that feels easy is the one to check, because its questions were probably copied from headings.
• Aim goes before the main learning event when it can, on Sunday for the week’s material; ten minutes before a class buys the largest chunks only, and after the class still works. The bar for the stage is a concept list that turns into questions finding chunks and the relationships between them, every time.

The two questions

Two questions carry the whole pass: why is this important? and how is this related to another concept? The first pushes toward purpose, function, and priority. It holds attention on the big picture while the detail goes past. The second forces comparison: dependency, contrast, cause and effect, and the relationships that were not obvious. It exists because the first question cannot get outside what is already believed about the material. Importance gets answered through the current view of the topic, and the link between two ideas often surfaces only when relating them is forced. Forcing a relationship between two things that seem unrelated is how the view changes. The pair is a working restriction, and it has never been measured as an optimum. It stands in for nearly every other question worth asking, and it is kept to two because a wide question set scatters the thinking and raises the odds of drifting far from the topic.

Relational questions are the ones worth this effort because the brain keeps what it can connect and drops what stands alone. A concept that many routes run through gets maintained. An isolated fact gets dropped, however long it was stared at, so effort spent building connections makes the individual facts stick as a by-product. A “what is this” question stays shallow for the same reason: it can be answered without connecting anything, and that lack of connection is the defect. A simple question is no defect in itself. Comparison also does the lower-order work by itself. Two things cannot be compared without learning what they are, and each definition arrives inside the comparison with its purpose attached. Definitions learned first and related later mostly do not survive, because nothing was integrated while the definition was still in mind. A definition learned on its own is mostly gone within minutes.

Answering either question takes a judgment of what matters and why. It is never a retrieval. The effort of that judgment is the work itself, and a connection only counts when the relationship it records means something. A map that is only connected is not yet coherent.

Before you start

Thirteen things should already be in place before the first Aim. You have felt that the quality of learning depends on the kind of thinking used. You have compared and contrasted instead of learning things one at a time. You have made non-linear notes, and you have felt how the way notes are written helps or blocks that thinking. You have used the traffic light drill, writing questions before reading and answering them after, and you have noticed that different questions open different relationships. You have formed backbones and seen a topic in a different order from the one it was taught in. You have run spaced interleaved retrieval. You have run Kolb’s cycle on your own attempts, asking what happened, what rule, and what changes next time, and you have felt how the depth of a reflection decides how useful the next experiment is. You count progress as marginal gains. You keep your time and tasks in order. If any of these is missing, the rule is to practise the current skills more before moving on. Without them the pass overloads, and the answer to that is more practice and never a smaller Aim. Moderate to high conscious competence in the earlier techniques is what prevents technique errors and overload.

Three of the thirteen are the drills the step rests on. The traffic light drill trains inquiry. Red light means write questions before you read. Green light means answer them. It is a drill, and once the questioning runs on its own the drill is dropped; run it until the questions come without it. Order control trains you to learn in your own order: take one chapter and learn it in your own order, jumping where the questions lead. The order a resource uses is a stranger’s guess about what you already know, and you almost always rearrange it, a little or a lot. Biggest chunks first trains the first half hour of a session. In that half hour the only goal is the three to six main chunks of the whole topic and how they relate: give a whole topic 20 to 30 minutes and stop with three to six chunks and how they relate. Hundreds of pages can be scoped that way in 15 to 30 minutes. Aim folds these three into one pass and raises the standard on the questions. It replaces the red light of the drill and replaces prestudy basics. The green light stays until Shoot replaces it.

Running the pass

The pass starts with a skim across every resource at once: video, book, slides, course guide, textbook, and a search engine or a model for what the resources lack. The skim covers headings, learning objectives, diagrams, summary boxes, keyword lists, lecture slides, problem sets, and chapter introductions. The keywords come off that skim deliberately out of the material’s order, so that the resource’s headings do not become the chunks. Keywords the material never names count too: a paragraph whose concept goes unnamed, a diagram carrying an idea with no label, each gets a keyword of the learner’s own making. For a term that cannot even be guessed at, a one-paragraph gloss from a search is enough, and then the pass moves on; that is scoping rather than studying. The list stays rough. It is raw material for the questions, and it is never a set of notes.

These sizes are the system’s working numbers, and none of them is a measurement. A topic the size of a couple of textbook chapters or a few weeks of class holds 15 to 30 major concepts: terms, processes, people, cases, formulas, mechanisms, arguments. Catching roughly 80 percent of them in about 25 minutes beats catching all of them in two hours. Completeness is not this step’s job. When a collection step runs long, one of three things is wrong: too many words are being collected, the volume is too high for the current skill, or detail is pulling the learner in. The collection itself is the easy part and can be handed to a transcription tool or a model. The thinking that follows cannot be handed off.

Then the two questions go over the list: why is this important, and how does this relate to that other concept. Top-down at first, because the earliest keywords have nothing to relate to yet. By the third or fourth there is enough on the table for the relating to work, and from then on the learner takes whichever keyword curiosity points to next. The list was collected out of order to begin with, so nothing is lost by jumping. No question asks what something is. Each keyword gets placed as it is worked, so the load never piles up. As the questions cluster, the concepts group into rough chunks: a shared function, a common cause, a sequence, a contrast, a problem-solution pair, a recurring mechanism, a theme, a set of examples serving one purpose. Importance-Based Chunking is what the why-important question is feeding: grouping by why things matter rather than by where they appeared. Each relationship gets its reason written on it. A line between two terms with no reason on it records nothing about why they are related; the written reason is the judgment.

The first structure is a hypothesis, and it is kept loose on purpose. Effort already spent on a structure makes the structure feel too expensive to break, and committing early is the sunk cost trap. In progress, the sketch looks like this: concept A seems important through B. C is made by A and B and feeds back to B. E leads to D. Later, maybe A and C belong together and lead to B. That collection of half-held groupings is what a real Aim looks like. Once rough chunks exist, a second question pass runs over the groups themselves, asking how the chunks affect each other, and the clean relations between the biggest ones are the backbone. The pass ends with a rough backbone of the main chunks and no details on it, and with a short list: questions to answer, relationships to verify, chunks to test, unclear terms to watch for. Shoot takes that list next: the working pass that hunts the material for answers and builds the map.

Order control

A resource’s order is a sequence somebody else made from a network. Text, slides, and video are linear because delivery demands it: the person who wrote the resource held the knowledge as a network and pressed it into a sequence to transmit it. Rebuilding the network is the learner’s job, and collecting keywords in the delivered order brings that sequence back in. The headings become the chunk structure without anyone deciding that, and the questions take on the writer’s frame instead of testing it. That is why the collection step jumps around; Order Control holds the general principle this pass applies in its first five minutes. Encoding format and retrieval format do not need to match. Material encoded in the structure that holds it can still be practiced in the linear format an exam will demand, so the order taken back during encoding costs nothing at retrieval.

Questions that work and questions that do not

Working questions usually start with why, how, when, what changes, or what depends on what. Ten stems cover most of the ground:

• Why does this matter?
• What problem does it solve?
• What would break without it?
• How does it relate to that other concept?
• What is the most important difference between these two?
• What causes it?
• What does it cause?
• Which examples belong together?
• What is this optimizing for?
• What would an examiner require of it?

Weak questions ask only for definitions: what is this, what does this word mean, what did the chapter say. Definitions are sometimes necessary and should not dominate the pass. The comparison work collects them anyway, purpose attached, so “what is it” has a place for rough orientation and never for memorizing. Anything answerable directly, without comparison or judgment, tends to encode weakly, and a question list built that way is one of the most common causes of weak chunking and disorganized maps.

Weak:

What is mitosis? What does "diploid" mean? What did the chapter say about checkpoints?

Strong:

What would break without checkpoints, and what is mitosis optimizing for that meiosis is not?

One failure has a shape worth recognizing: the repackaged answer. A fact already held, plants make energy by photosynthesis, gets inverted into “how do plants make energy?”, and the answer is the fact it started as. It looks like a question and it finds no gap. The selection rule that prevents it: only questions that cannot already be answered. Every such question lands on a weak point, so an hour spent answering them goes entirely to weaknesses. An hour of ready-made questions may reach one or two gaps and leaves the others unfound. A short session that produces nothing but correct answers is evidence the gaps have not been found yet. Relevance selects too: material that is personally live, or already interesting, gets connected more willingly because the learner already has a reason to care about it, and a question whose answer is already in hand cannot supply one.

When to aim, and the smallest version that works

The pass does the most good before the main learning event: before a lecture, before the reading, before a video, before a dense self-study block, before revising a topic that feels fragmented. In the standard week, Aim goes on Sunday for the coming week’s material. Prestudy is the timing layer beside the system. It decides when learning starts and builds the rough frame, and Aim sharpens that frame into questions. Prestudy has no step of its own inside the system: prestudy is when learning starts, the Bear Hunter System is how it happens, and Aim is the first step. When Aim happens before the class that covers the topic, it is prestudy. How long the prestudy runs is a decision. Five or ten minutes buys a superficial Aim, the largest chunks and the major relationships only, and even that pays. More time buys a full Aim and part of Shoot. When the class is compulsory, studying everything beforehand wastes the hours you have to sit there anyway. Arrive primed so the class counts, and finish the Aim after it.

When before is impossible, immediately after still works. It is worse than aiming before, and it is far better than never aiming. Aim after the class with no prestudy is the messiest order and never the one to plan for. The notes made in class become one more resource for the Aim, and they already carry some grouping, so the pass runs a little shorter. Doing the Aim inside the class itself is hard, because that kind of thinking does not fit an intensive lecture.

The smallest version that works, in the same working sizes:

1. A list of five to ten key terms.
2. The three that matter most, picked out.
3. “Why is this important?” asked of each of the three.
4. “How do these relate to each other?” asked across them.
5. One rough backbone: the three to seven biggest chunks.

Five to ten minutes covers it. It is a cut-down of the full pass and does not replace it, and it is the default opening move on any new topic.

The pass is the first part of learning

Writing the questions is already learning. The judging and comparing they demand is the thinking the rest of the loop runs on. Progress gets measured on what can be recalled and used, never on pages covered or notes produced.

The cost comes at the start. First passes are slow, because every major concept takes both questions and both answers take thought. The cost falls as the structure already built starts absorbing new material: the next concept is recognizably like one already placed, so later topics attach faster than early ones. The slow phase is where most learners stop. A learner who stops there has paid the full cost of the slow phase and gets none of the speed-up after it. Skill raises the tolerance for the load, and the effort stays the same: the load holds where it is, and handling it stops being uncomfortable. That is the sense in which it gets easier.

The pass also has prerequisites. It assumes comfort with grouping ideas, judging what matters, and making notes non-linearly, all three at once; without them it overloads, and they are worth practicing separately first. Scope runs wide and shallow: the largest topic that can be tolerated in one pass, usually a whole topic. A shallow pass over a whole topic, with gaps in it, still collects most of the major concepts. A detailed pass over one small part of the topic collects almost none of them, wherever it lands. As skill grows, one pass covers a whole topic. The boundary bends outward a little too: a few minutes on something adjacent that unlocks the material is usually cheaper than the re-learning caused by staying narrow, while far outside the topic is the drift the two-question restriction exists to prevent.

How it feels

The pass is expected to feel slow, effortful, and uncertain. Going well, it produces curiosity about the answers, relationships that had not been visible before, a growing suspicion that the material’s order is not the best one, and a mild overwhelm at how many groupings are possible. The strongest sign is a question whose answer is wanted for its own sake. These feelings are how the pass is checked: Self-Regulation is the control layer that judges a technique by the thinking it produces, and during this pass the thinking can be felt while it happens.

The pass is meant to feel bad in specific ways, and each feeling has a meaning. Chaotic, from going back and forth between the questions and the material: that is integrative thinking, building the network. Confusing, while an answer is being worked out: that is deep processing. Overwhelmed, while a non-linear structure is being built: that is prioritising and evaluating, the highest order of the work. Curious and effortful: more of the brain is in use. Insecure, while details are skipped and the order is your own: that is the brain forming schemas in its own order and ignoring information until it becomes relevant.

Aim adds three feelings of its own. Overwhelmed by the number of relationships the why question raises: that is a network of possibilities waiting to be refined. Confused because many non-linear structures are possible and choosing takes effort: that choosing is the evaluation the pass exists for. Slow, because each keyword gets a deep look: the depth is what consolidates the learning, and the slowness fades as speed grows. The feeling to distrust is ease. A pass that feels easy is usually a pass where the questions were copied from headings. After the session, run Kolb’s cycle: what happened, what rule, what changes next time.

A pass that worked leaves this behind:

• More why and how on the list than what.
• The questions raise curiosity rather than duty.
• The topic’s overall importance got its own question.
• At least a few concepts got compared directly.
• Rough chunks exist, loosely held.
• The next pass has a target.

Where it goes wrong, and the repair

The weak pass is recognizable while it happens: questions that are only definitions, headings copied into question form, questions whose answers were already held, the uncomfortable question avoided, the structure polished before any learning starts. Its products carry the same signature: answers that are a single fact rather than a set of relationships, a list that is mostly “what,” why-important skipped for most concepts, questions written for concepts but never for the groups holding them, and maps that feel disorganized because the backbone was never built from importance and relationship questions. Unfamiliar exam questions that feel impossible usually point the same way, to concepts never related during encoding, though that reads as a sign rather than a law.

The repair that matters most is de-linearization. Stuck inside the material’s order, the headings convert into a flat list and the list re-groups by importance. That one move re-runs the main work of the pass on material whose order had taken over.

The pass produces	The repair
Too many questions	Only the questions that would change the map stay.
Questions too basic	”What is X?” becomes “why does X matter?”
No curiosity anywhere	The questions move to what confused, surprised, or looked useful.
Detail taking over	Attention returns to the biggest concepts and their relationships.
”Why important” ticked with one reason	The question re-runs until several relationships of different weight sit side by side.
Relationships multiplying into chaos	A better central chunk and a re-chunk, never more arrows.

Two failures have names. Importance-checklisting is “why is this important” answered with one reason and moved past: the question’s shape without its work, since most things matter for several reasons of different weight and the question has only finished when they can be compared. Spiderwebbing is relationships multiplying into chaos, and its repair lives on Importance-Based Chunking: a better central chunk, then a re-chunk. The warning that covers all of it: when the pass starts to feel like copying headings into question form, it has become a shortcut, the family The Shortcut Problem names, and it fails the test in The Technique Is Only as Good as the Thinking It Produces: the technique’s motions without the judgment those motions exist to force.

The working test runs over weeks. Within a few topics, new material in the same subject should attach faster and feel more familiar on first contact. If nothing attaches faster after several topics, the questions are almost certainly producing headings rather than relationships.

Run well, the pass takes a topic that arrived as an unsorted set of terms and leaves it as a small set of important questions, tensions, and probable groupings, and Shoot carries those questions into the material to hunt the answers down.

The bar

The bar for Aim is one sentence: you can consistently use concept lists to create high-quality questions that identify chunks, sub-chunks, and the relationships between them. Four other things are graded at the same point, because they have to keep running while Aim is learned. Focus is consistent and easy to set up in different environments. The green light of the drill still answers the Aim questions, in processed notes that show chunks and relationships. Spaced interleaved retrieval is near unconscious competence, with several interleaving methods and consistent spacing. Non-linear notes are becoming substantially more comfortable.

The bar is about the step, and on purpose it says nothing about the standard. Whether the chunks are importance-based, whether the reasons are root reasons, whether a map built from these questions has single-node chains or reverse causality: all of that is graded one stage later, when Shoot maps are read. Move on when the step comes out consistently; the chunks do not have to be perfect yet. That grading is the second half of Aim, on the page Aim - Training II.

How to practice this

1. Before the main reading, skim every resource for keywords, out of order, for about 25 minutes. Notice whether the collection runs long. If it does, too many words, too much volume, or detail is the cause.
2. Ask two questions of each keyword: why is this important, and how does it relate to that other concept. Notice whether any question could already be answered; that one finds no gap.
3. Sketch rough chunks, hold them loosely, then ask how the chunks affect each other. Notice any urge to keep a structure because of the effort already spent on it.
4. Stop with a backbone of the three to seven biggest chunks and no details on it. Notice whether the list has more why and how on it than what.
5. Check how the pass felt. Chaotic, confusing, overwhelmed, and slow are expected. If it felt easy, check whether the questions were copied from headings.
6. After the session, run Kolb’s cycle: what happened, what rule, what changes next time. Over several topics, notice whether new material in the same subject attaches faster on first contact.

Related pages

• Deep Processing: the parent dimension; Aim is one encoding operator inside it.
• Bear Hunter System: the parent loop and its three schedules; Aim is its first pass, ahead of the working and pruning passes.
• Shoot: the next pass; the material worked against Aim’s questions into a rough map.
• Skin: the later pass; the map pruned to what survives without the material; it plays no part during Aim.
• Prestudy: the timing layer that builds the rough frame before the main event; it decides when learning starts and has no step of its own inside the system.

7 more related pages

• Importance-Based Chunking: where the why-important question leads; grouping by why things matter, the home of the spiderwebbing repair, and the standard Aim - Training II holds the step to.
• Order Control: the general principle behind out-of-order collection; the rule that the order is yours.
• Inquiry-Based Learning: holds the traffic light drill.
• Kolbs Experiential Cycle: the check that runs after every session.
• Self-Regulation: the control layer; a technique’s quality is judged by the thinking it produces.
• The Technique Is Only as Good as the Thinking It Produces: the quality test that heading-copying fails.
• The Shortcut Problem: the family the heading-copy failure belongs to.

Open questions

• The readiness list is thirteen items long and none of them comes with a test. What is the shortest honest self-check for each one?
• How many topics does it take before the 25-minute session holds at 80 percent coverage?

Sources

1. Pressley, McDaniel, Turnure, Wood & Ahmad (1987), Journal of Experimental Psychology: General. Elaborative interrogation: answering “why” questions improves factual memory.
2. Dunlosky, Rawson, Marsh, Nathan & Willingham (2013), Psychological Science in the Public Interest 14(1). Elaborative interrogation and self-explanation rated moderate utility.
3. Craik & Lockhart (1972), Journal of Verbal Learning and Verbal Behavior 11, 671–684, and Craik & Tulving (1975), Journal of Experimental Psychology: General. Semantic encoding outlasts structural encoding.
4. Chi, de Leeuw, Chiu & LaVancher (1994). Self-explanation and relational elaboration.
5. Bjork (1994). Desirable difficulties: the effort of encoding is the work.
6. Barnett & Ceci (2002). The transfer literature, which is why relational-encoding failure on unfamiliar exams ships as a sign rather than a law.
7. The iCanStudy program by Justin Sung, learning.icanstudy.com. Ascent I: Bear Hunter System: Aim, How to Aim, Ascent I Check, and the sub-lesson Common chunking mistakes. Technique Training: Inquiry-based learning, Doing TLS correctly, Order control. Ascent II, Barriers for shooting, for the line that ease is the alarm. Private course text captured to the vault’s raw folder, never republished, read in full on 2026-09-07. Every working number on this page is the program’s own, and none of them is a measurement.